Stability Enhancement of Inverter-dominated power systems with virtual inertia control by Lalitha Subramanian -THÈSE Pour obtenir le grade de DOCTEUR DE L’UNIVERSITÉ GRENOBLE ALPES Spécialité : GENIE ELECTRIQUE

Stability Enhancement of Inverter-dominated power systems with virtual inertia control Lalitha Subramanian THÈSE Pour obtenir le grade de DOCTEUR DE L’UNIVERSITÉ GRENOBLE ALPES Spécialité : GENIE ELECTRIQUE 

 SUMMARY

 The electric power system has been traditionally energized by synchronous machines like steam turbines, hydro turbines, and diesel engines. These rotating machines inherently contribute to the system resilience by providing rotational inertia. The presence of an adequate inertia in the system provides the liberty of allowing a control delay for the governor-input valve controls to respond to the frequency deviation. With the displacement of synchronous machines by converter-connected sources, the reduction of inherent system inertia is evident. However, there is also a counterpoised observation that the required amount of inertia in the transformed power system is reduced, given the faster response of the converter-based DERs. Therefore, we resort to synthetic inertia to improve the resilience of a low-inertia grid. In this context, this thesis explores questions such as: What is the adequate synthetic inertia/frequency response capability for a stable power system? How can we quantify the flexibility required to provide this adequate inertia? Does inertia greater than the adequate level necessarily indicate a higher stability margin? How different is the effect of distributed synthetic inertia on the oscillatory stability compared to synchronous inertia? Firstly, the aspects of flexibility and methods to characterize them for an adequate synthetic inertia and fast-frequency response are addressed. A generalized virtual storage flexibility model has been proposed to quantify the heterogeneous bidirectional flexibilities and their combination to provide a certain level of synthetic inertia. As an illustration, a hybrid energy storage system has been sized to provide synthetic inertia and fast-frequency response for a standard power network.The subsequent chapters discuss synthetic inertia and fast-frequency control actuated by PV systems with hybrid energy storage. In this thesis, inverter control has been explored with a complete DC-side model takes into account the effects of PV intermittency, unlike most research works on inverter control that assumes a sufficiently large DC source/sink. Synthetic inertia controllers are categorized as grid-following and grid-forming topologies, which significantly affect their impact on system stability. Conventionally, the inertia and damping parameters are tuned and fixed over a scheduled time slot based on the available flexibility. It has been identified that a higher inertia is required on the occurrence of a disturbance to limit the rate of frequency deviation and a higher damping is required for a faster settling time. Therefore, for each of the control topologies, a rule-based real-time inertia tuner has been proposed to optimize the frequency deviation, its rate, and the settling time. The algorithm has been improved through a model predictive control with a rate-based linearization. The rate-based linearization extends the model validity to the transient zones.For systems with multiple grid-formers and multiple frequency responsive units, a distributed optimization problem has been formulated and solved to collectively tune the inertia and damping parameters which are constrained by the available flexibilities.The efficacy of distributed grid-forming and grid-following synthetic inertia in replacing their synchronous counterpart in a microgrid has been compared. Microgrid regulation in grid-connected and islanded modes have been studied by modeling the DERs with discussed control strategies. The impact of the two types of synthetic inertia controls on the small signal stability of the system are examined by modal analysis and bifurcation plots to derive the conditions for oscillatory stability in a microgrid with distributed synthetic inertia reserves. The effectiveness of the proposed control strategies in restoring the frequency stability of low-inertia systems has been validated by power hardware-in-the-loop experimentation.

VIEW FULL THESIS: https://theses.hal.science/tel-03505925v1/document

Multilevel Multiplexed Inverters for Applications up to 1500 Volts and 100 kW. by Kepa Odriozola Sagasta

Multilevel Multiplexed Inverters for Applications up to 1500 Volts and 100 kW. by Kepa Odriozola Sagasta

SUMMARY

 Renewable energy, including solar and wind, is at the heart of the transition to a less carbon intensive and more sustainable energy system. Over the past two decades, there has been a significant increase in photovoltaic (PV) power installed in new solar power plants. This increase in power has also led to an increase in DC bus voltage up to 1500V (the limit between LV and MV in DC). It is worth noting that grid integration of distributed energy sources such as PV systems with battery based energy storage systems (BESS), both utility-scale and residential and industrial, is gaining importance. At the same time, the global data sphere is projected to grow from 33 zettabytes (ZB, 1021 bytes, one trillion gigabytes) in 2018 to 175 ZB in 2025. Due to the data tsunami, digitization, high-speed wireless networks, new data-intensive technologies, and the growing demand for cloud-computing have led to the development of data centers, which have become an electricity intensive industry. For example, several studies have estimated that between 2025 and 2050, the ICT industry could account for up to 20% of global electricity consumption. The increase in installed power of data centers and solar power plants implies high power converters (MW or more) connected in parallel. The increase in voltage for PV and energy storage is a strong challenge to reduce the level of current in the devices in order to obtain more compact systems, because today, with the current solutions, the objectives of efficiency, cost and power density are not met. Therefore, new high-efficiency power electronic systems and associated control strategies are needed to make this transition possible. This thesis is part of the study and design of a new family of high efficiency three phase multilevel converter topologies, called Multiplexed, intended for UPS, Energy Storage and Solar Inverter applications. The case studied, foresees a converter with a high voltage DC bus with a voltage variation range from 900VDC to 1500VDC, connected to a grid or a three-phase load of American (480VAC/60Hz) or European (400VAC/50Hz) voltage, and a bidirectional power ranging from a few tens of kW to 100kW. First, we will explain the general structure of multiplexed topologies. Basically, the structure is composed of a chopper stage (consisting of two symmetrical DC-DC converters) connected to a three phase DC-AC inverter stage. There are many variants depending on the topologies chosen for each conversion stage. In a second part, we will present the selected variants as a solution to the studied applications. We will also specify, the approach adopted for the dimensioning of these solutions, by considering the active elements (semiconductors) and the passive elements. We will present the performance results obtained via simulation. We will specify the control principle associated with this new range of topologies. Indeed,one of the main features of these structures is that the two stages, chopper and inverter, are connected without intermediate filter which implies that the inverter must re-switch the voltage that has already been switched by the chopper stage. The PWM modulation strategy that has been developed for this purpose will then be presented and detailed. Finally, based on the models and results obtained in the previous sections, we will show the realization of the 100kW prototypes built in order to validate the architecture and the control strategy proposed around the 'Multiplexed' concept.

VIEW FULL THESIS: https://theses.hal.science/tel-04222473v1/document

The study of the correlation between the microstructural properties and the electrochemical performance of the Li-ion battery high-energy-density positive electrode Tuan-Tu Nguyen

 

SUMMARY 

Although “post” Li-ion battery is arising as an inevitable solution for sustainable energy transition, it would be unwise to assume 'conventional' Li-ion battery is approaching the end of their era; as many strategies are still available to improve their performance. While progress has been continuously achieved to get even better active materials, industry engineers and academic researchers have kept improving on the electrode scale. The most direct way can be done through the microstructure design. In light of this, one attempts to understand the interplay between the electrode microstructure and its performance in this work, which plays a vital role in achieving high-performance Li-ion battery electrodes. This work relates to three major pillars, which are electrochemical measurements, tomography and numerical modeling. The LiNi0.5Mn0.3Co0.2O2 industry-grade electrodes are investigated, since LiNixMnyCo(1-xy)O2 materials constitute a popular class of cathode materials. The first part of this work focuses on the first two pillars that allow a complete characterization of porous electrodes, including both electrochemical performance and microstructural properties. Appropriate experimental methods are carried out to determine the transport properties of the electrodes, such as electrode tortuosity factor and effective electronic conductivity. Thin electrodes are made for the determination of active-material intrinsic properties. The performance of industry-grade electrodes is then assessed through discharge rate capability. A complete quantitative analysis of the microstructures of industry-grade electrodes using the X-ray holotomography technique is performed. The microstructural heterogeneities are quantified for each phase (active materials, carbon binder domain, pore space) separately, along with the statistical quantification of their inter-connectivity at the particle scale. Besides, Operando X-ray Absorption Near Edge Structure coupled with transmission X-ray nano computed tomography are done, offering a direct correlation between electrode microstructure and local electrochemical performance. Also, an image quality assessment method is investigated, which utilizes convolutional neural networks. It can be a direct tool to produce reliable segmentation results and guide the image pre-processing step (eg denoise, contrast enhancement) for quality enhancement. The second part relies on the numerical approach to further understand the underlying physics of the electrode during operation. One starts with introducing a new concept of the tortuosity factor, which is demonstrated through a numerical approach to be more appropriate for porous electrodes. A model representing the symmetric cell method is implemented in an open-source application called TauFactor for the electrode tortuosity factor determination using tomographic data. Then,the performance of four industry-grade electrodes is investigated through mathematical models. The parameterization of the models is carried out carefully using appropriate experimental methods. As the Newman pseudo-2D model fails to capture the behavior of the set of electrodes, the formation of porous agglomerates due to the calendering process to achieve high-energy-density is identified to be responsible for this discrepancy. Thus, porous agglomerates are included in the Newman pseudo-2D model. The validation of the electrodes with different electrolytes is done. As a result, the porous agglomerate effects are identified as a dominant limiting factor at high C-rates for high-energy-density electrodes.

VIEW FULL THESIS :https://theses.hal.science/tel-03989879v1/document

Multi-physical modeling of the graphite electrode within a lithium-ion battery: Study of heterogeneities and aging mechanisms Nicolas Dufour

 SUMMARY

The negative electrode of lithium-ion batteries is commonly made of graphite. Although having an interesting specific capacity, aging, intercalation kinetics and lithium transport both in the active material and the electrode porosities limit its optimal and homogeneous operation. In this thesis work, the mechanisms behind these limits are explained using a multi-physics model of the porous electrode type. A sensitivity study of the model parameters showed the importance of the parameters related to intercalation kinetics and lithium transport in solid and liquid phases. The exploitation of the model, validated experimentally, shows that, during the operation of the electrode, the appearances of lithiation heterogeneity are correlated with the particular shape of the equilibrium potential of graphite with respect to its lithiation rate. The modeling of the particle size distribution greatly amplifies these heterogeneities and significantly degrades the overall performance of the electrode. As a first approach, an operando measurement of the distribution of lithiation states confirms the heterogeneous aspect of the electrode operation. Data on the cycling and calendar performances of graphite-NMC cells have allowed the construction of different electrode aging models. The growth of the passivation layer (SEI) can alone explain the loss of cyclable lithium. The SEI heterogeneities obtained by the model are negligible in the current state. The capacity gains and sudden losses are explained respectively by SEI dissolution and lithium-plating formation mechanisms.

VIEW FULL THESIS: https://theses.hal.science/tel-02148211v1/document

100kW, 10kHz Medium Frequency Transformer Design Optimization and Experimental Verification Marko Mogorovic, Student Member, IEEE, Drazen Dujic, Senior Member, IEEE

 100kW, 10kHz Medium Frequency Transformer Design Optimization and Experimental Verification Marko Mogorovic, Student Member, IEEE, Drazen Dujic, Senior Member, IEEE 

 Abstract—This paper describes a novel, model based, design optimization methodology for high-power medium frequency transformers for medium-voltage high-power electronic applications, namely emerging solid state transformers. Presented procedure enables the designer to interactively select the most optimal design in a simple and intuitive way, while inherently offering flexibility in terms of various design alternatives, depending on the component availability. The core of the design algorithm is explained in detail together with all of the relevant modeling and technical challenges associated with realization of a prototype. A 100 kW, 10 kHz medium frequency transformer prototype has been designed according to the presented algorithm, practically realized and thoroughly tested in order to verify the theoretical design.

VIEW FULL PAPER :

 https://infoscience.epfl.ch/entities/publication/7d2c7186-c374-4bda-a35a-88cd849f908c